IT&Software

Instagram now uses AI to block offensive comments

Instagram now uses AI to block offensive comments

The Facebook-owned platform said the AI filters have been "trained" to identify and block comments from appearing underneath posts if they are recognised as either offensive or spam-related.

Instagram has announced two new tools that use artificial intelligence (AI) to block offensive and spam comments on posts to the photo sharing app.

A button is being added to the app's settings under the Comments section, enabling users to turn on the comment blocker.

Instagram says both types of filters are powered by machine learning and can of course improve over time as they see more of the types of content that should be blocked.

People can choose to have the automatic comment filter on or off. Instagram says it's the next step in its commitment to "foster kind, inclusive communities", but really it's just the right thing to do for our collective sanity online.

More news: Kylie & Kendall Jenner Ripped For Disrespecting Notorious BIG & 2Pac's Legacy?
More news: Rain chances move out and highs warm up
More news: House panel wants to interview longtime Trump bodyguard in Russian Federation probe

The spam filter, which is separate from the mean comments tool and being applied automatically, is available in English, Spanish, Portuguese, Arabic, French, German, Russian, Japanese and Chinese.

The spam filter looks for any obvious spam in comments, blocking it from your posts and live videos.

For instance, it would not block sentences which contain words like "black" or "brown", but would be able to tell when they were being used in a racist context. After doing so, Instagram will automatically hide potentially unsettling comments on your posts. They then fed most of this data into DeepText, which created algorithms based on what it found in the spam comments.

While it's unclear exactly how well DeepText worked in this trial run, Wired reports that Instagram CEO Kevin Systrom was "delighted".

This new system in filtering out hateful and spammy comments might not work perfectly at first, but it should improve over time.